OF DWARF ELEPHANTGRASS UNDER GRAZING' LUIS ROBERTO DE A. RODRIGUES2, GERALO O. MOTT3, JONAS B. VEIGA4
eWILLIAM R. OCUMPAUGH5
ABSTRACT - Morphological chaxacteristics and tilleiing of dwarf elephantgrass (Penn&etum
purpu-reum Schum.) weie studied in a grazing trial conducted at the University aí Florida - USA. Tise grass was subjected to fite effects oí two management factors, each at five leveis as foliows: length oí graring cycle (GC) (continuous grazing, 14,28,42 and 56 days) and grazing pressure (GP) (500, 1,000, 1,500, 2,000 and 2,500 kg aí residual leaf dry matter ha'). Response suríace methodology was used to analyse the data. Increases in CC and decreases ia GP increased plant heigbt, apícai meristems heíght, number aí internadas and dry matter produced per tiller. The percentage oí apicai meristems dm1-nated was greater under conlinuous grazing and high CI'. The number aí axiilary buds per tiller was depressed under high CI'. Shorter internodes were observed at short GC and higb CP. Paimary tillers constituted the predominant íorm of tiliering in this grass. The results obtained indicate that short grazing cycles and high grazing pressures should be avoided in the management aí dwarfelephantgrass ia arder to insure persistence and productivity oí the pasture.
Index terms:Fennisezum purpureurn, grazing trlal, pasture management, management factoas PERFILHAMENTO E CARACTERÍSTICAS MORFOLÓGICAS
DO CAPIM-ELEFANTE ANÃO SOB PASTEJO
RESUMO . O perfilharnento e as características morfológicas do capim-elefante an5o (Pernisetum
par-pureum Schumj foram estudados num ensaio de pastejo conduzido na Universidade da Flórida - EUA. O capim foi submetido a 5 níveis de presso de pastejo - PP (500, 1.000, 1.500, 2.000 e 2.500 kg de MS de folha residual tia1) e ciclo de pastejo - CP (pastejo contínuo, 14,28,42 e 56 dias). Os dados foram analisados através da metodologia de superfície de resposta. Aumentos no CI' e reduções na PP resultaram no aumento da altura das plantas e dos meristemas apicais, no número de entrenás por perfilho e na produç3o de MS produzida por perfilho. A percentagem de meristemas apicais eliminados foi mais elevada sob pastejo contínuo e alta PP. O número de gemas axilares por perfilho foi reduzido sot, alta PP. Entrenós mais curtos foram observados em CP curtos e alta PP. Perfilhos primários consti-tuem a forma predominante de perfilharnento neste capim. Os resultados obtidas sugerem que CP cur-tos e altas PP devem ser evitados no manejo do capim-elefante anuo a fim de assegurar a persistência e a produtividade da pastagem.
Termos para indexaç5o: Pennisetum purpureum, ensaio de pastejo, manejo de pastagens, fatores de manejo.
INTRODUCTION
Elephantgrass (Pennisetum purpureum Schum.) is valued for its herbage yields, palatability, and
Accepted for publication on August 22, 1986. Part aí the dissertation presented by the tlrst author to the University aí Florida in parfial fulfiilment aí the requirements for the degree aí Doctor aí Phuloso-phy.
2 Assistant Professor, PhD., FCAVJ/UNESP, Caixa Pos-tal 188, CEP 14870 Jaboficabal, SP, Brazil.
Professor aí Agronomy, PhD., bniversity aí Florida, Cainesvifle, FL., 32611, USA, Deceased, November 26,1984. -
Eng. - Agr., Ph.D.: EMBRAPA/CPATU. Caixa Postal 48, CEP 66000 BeWm,PA, Brazil.
$ Associate Professor, Ph.D., Texas A & M University. Coliege Stafion,TX 77840, USA.
good quality herbage (Bogdan 1977). Although widely culdvated as a fodder plant, it is not com-monly used in pastures. Many researchers have suggested that the main limitation af elephantgrass under grazing results from management problems deriving from its growth habit. Differences in habit among grasses are due to the pattern aí lateral shoot development and the length aí stem inter-nades (,Jewiss 1966). The elongation afinternodes reduces the density af buds dose to the ground, elevates the canopy, and increase apical meristems vulnerability to cutting ar grazing. Thus, the pasitian aí the growing paints at the time aí cut-ting ar grazing is an iinportant factor in deter-mining suscepübiity af forage plants ta defoliation
Gomide etal. 1979, Doyrat etal. 1980, Humphreys 1981).
Dahl & 1-lyder (1977) indicated that tillering ability from axilary buds without the apical meristem being removed is considered an indication of an efficient forage producer. During growth of a controlled sward, the dry matter yield of shoots increases because of both the weight increase of individual tillers and the increase iii number of tiiers (Humphreys 1966, 1981,Dovratetal. 1980). However, the rate of tilier initiadon lii the pasture environment may be significandy affected y grazing.
Youngner (1972) concluded that, when defoli-ation removes apices of elongated stems, apical dominance is broken and the rate of tilier produc-tion is increased. On the other hand, the intensity of defoliation cnn affect the characteristics ofnew tiliers. Thus, Belyuchenko (1980) observed that the height of cutüng modified the proportion of new tillers in the regrowth ofelephantgrass. When planta were cut dose to the ground, tillers ori& -nated from buds on the rhizome. Cuts at 10 or 15 cm produced an appreciable number of basal buda arising from the crown tillering area. Cutting at 15 cm or higher led to a predominance of aerial tiiers arising from axillary buds on the stems.
Recently, a dwarf type of elephantgrass, known as lime N-75, was isolated from a population ofP.
purpureum in the Constal Plains Experimenta! Station, Tiftom, Geora. In 1979 this material was introduced at the Beef Research Únit of the Uni-versity of Florida where it has been studied (Bod-dorff 1982, Castillo-Gallegos 1983, Veiga 1983, Rodrigues 1984). The objective of this research was to study the effects of grazing pressure and length of grazing cyc!e om morphological charac-teristics and tillering of dwarf elephantgrass under grazing.
MATERIAL AND METHODS
A grazing experiment was conducted, from April to November 1982, at the Beef Research Unit of tbe Umi-versity of Florida, Gainesvifle, to study morphological and physiological responses of dwarf elephantgrass
(Pennise-lum purpureum Schum.) under grazing conditions. The experimental pastures were established with dwarf ele-phantgrass, lime N-75, introduced from Dr. W.W. Hanna's nursery, located at Tifton, Georgia. The soU of the ama
belongs to Oiipley sejies (Order Entisol), with Pomona series (Order Spodosol) occurring on a smaU part of the site. The climate is subtropical and humid with an average annual precipitation of 1,300 mm and a frost-free period of 276 days. lhe dwarf elephantgrass was propagated vegetatively by both small rooted cuttingsand stem Gut-tinga of 5 cm to 10 cm, planted iii 1 m x 1 m spacing. lhe
plantng of the grass began in the summer of 1980 and was concluded in July 1981. Irrigation, weed control, and a complete fertilizer with micronutrients were provided during the estabiishment phase to insure a good stand of plants. In order to test the methodology, a prehminary grazing study was conducted from August to October 1981. Maintenance fertilizei was applied during 1981 and at the beginning of the 1982 grazing season.
Two grazing management factors were studied as experimental vatiables: a) length of grazing cycle, and b) grazing pressurç. lhe grass was subjected to five leveis of each of these variables as follows: length of grazing cycle: continuous grazing, 14, 28, 42, and 56 days, and grazing pressure: 500, 1,000, 1,500, 2,000 and 2.500kg of residual leaf dry matter (RLDM hí t ). Each grazing cycle, except for the continuously grazed treatments, consisted of two days of grazing followed by an appropxi-ate rest period to complete the desired length of grazing cycle. Because residual dry matter is subjected to measure-ment errora, ali efforts were made to have RLDM ha 1 as dose as possible to the projected values. However, tluc-tuations in these values were observed throughout the grazing season and for clarity figures yreseted in this papei are extended to 250kg RLDM hi
Response surface methodology was used to study 13 treatment combinations, arranged in a modified nora-rotatable central composite design according to Littell & Mott (1975). Ali the design points shown in Fig. 1 were replicated tMce, and the resulüng 26 experimental units were disttibuted in a completely randomized design in the tield. lhe central treatment (1,500kg RLDM and 28-day grazing cycie) was selected to represent the region where the combination of experimental vatiables was anticipated to be near optimum lo insure produc-tivity, quality, and persistence of lhe grass. Tire data obtained were analyzed using lhe complete second order polynomial model, Y & b 0 + b1x 1 + b 1 x2 + b 11 x 1 2 + + b 22 x 2 2 + b 12 x 1 x 2 + e, whem y observed response, x l -gruingcyclo,mdX2'estiTnatedgrazingpmssum.
Grazing pressure was imposed as residual leaf dry matter in kg ha t and was estimated byVeiga (1983) using a double sampling technique. Visual estimation was used to determine tire levei of grazing pressure attained during the grazing period. II should be indicated that a lenient grazing pressure suggests a large amount of residual leaf left at the end of tire grazing period and conversely a high grazing pressure suggests a small amount of residual leaL Stocking iates were detemrined on tire basis of available forage, and lhe number of animais was adjusted during tire grazing petiod in order to achieve tire desired Pesq. agropec. bras., Brasília, 21(ll):1209-1218, nov. 1986.
cm 70 46
\ \
2500 'o -c 2000 o —J CP 500 UJ D ci) (1) w 1000 o z Ni 50C grazing pressure in about two days. Animais were usedsolely lo defoliate lhe pastures since the experiment was planned-to study the effect aí lhe animais on the pasture regrowth and not the effect of lhe pasture on lhe animais.
GRAZING CYCLE (X1) -2-2 0,-2 t2,-2 -1,-! 9 LIJI DI
g
k-2,01
0
Y
z 1 ct O 1 Ri (27 -2~2 0+2 1 ~2+2• FACTORIAL PO1NTS 22 * CORRER POINTS 22
OAXIAL POINTS (2)121 *CENTER POLNT
FIG, 1. Configuration of the rnodified central composite non-rotatable design, including lhe codod treat-ment combinations.
The study oí lhe morphology of dwarf elephantgrass consisted in lhe measurements of the foliowing morpho-logical chaiacteaistics: apicai meiisterns height, stem height, number of axiliary buds and intemodes per tilier, and length of intemodes. For this purpose ten tilien were sampled at random before and after each grazing cycle, cut lo ground levei, and immediately taken to the labora-tory. Piant height was measured as the iength of the extended filiei from ground levei to lhe ligule of the last expanded leal. Percentage of apical meristems eli,ninated alter grazing was calcuiated in relafion to lhe total of tillers collected.
Thè number of tillers per plant was recorded before each grazing period in lhe rotationaily giazed pastures and at each 28-day interval in lhe conünuousiy grazed pas-tures. Two lillers were marked vdthin each ten plants chosen at randon in each pasture. Suitable tiliers were marked with hand made rings of plastic covered wire.
RESULTS AND DISCUSSION
Morphology of Dwarf Elephantgrass
Plant height, measured in terms of diler height, varied from 17.2 cm (250 kg RLDM ha' and
continuous grazing) to 104.1 cm (2,500 kg RLDM ha' and 56-day cycle). As the grazing cycle was increased and the grazing pressure was decreased, the plant height was increased (Figure 2). Increase in height as the plant advances in age has been reported in severa1 tropical grasses (Tardin et ai. 1971, Nascimento Júnior & Pinheiro 1975, Gomi-de et ai. 1979, Garcia & Silva 1980), including tall varieties of elephantgrass (Vieira & Gomide 1968, Pedreira & Boin 1969, Andrade & Gomide 1971).
250
O 14 29 42 56
GRAZING CYCLE (doys)
FIG. 2. Contour map of plant height of dwarf ele-phantgrass as affected by length of graring cycle
and grazing pressure (A2 = 0.98, CV 5.24%,
PC 0.01).
An analysis of Fig. 3 indicates that the apical meristems were reladvely '0W at high grazing
pressures and short grazing cycles and increased in height as the grazing pressure decreased and the grazing cycle increased. The response is similar to that of plant height suggesting that the grass tends to change its form under different grazing condi-dons. The data also suggest that grazing pressure had the greatest effect upon the elimination of apical meristems. The percentage of apical meri-stems eliminated varied frorn 25.0% (2,500 kg RLDM ha and 56-day cycle) to 73.3% (250 kg
RLDM ha and conúnuous grazing). It should be mentioned that at the be&nning of the growing period the apical meristems were dose to the ground in ali pastures. However, the apices of some stems were grazed in pastures subjected to heavier grazing pressure (250 kg RLDM hi'). On the other hand, in the second haif of the experi-mental period, after about August 1 st, lower stock-ing rates were required to attain the projected graz-ing pressure which may have reduced the number of apical meristems removed at heavy grazing pres-sures and short grazing cycles. In fact, the use of high stocking rates or the maintenance of animais in the continuously grazed pastures resulted in the elimination of a large number of apical meristems.
li has been suggested by severa1 authors that the position of the growing points at the time of defoliation is an important morphological factor in determining the persistence of forage grasses (Booysen et ai. 1963, Jewiss 1966, Langer 1972, Youngner 1972, Dali & Hyder 1977, Gomide et ai. 1979, Dovrat et ai. 1980, Humphreys 1981). Indeed, regrowth of tropical grasses is associated with the fate of apical meristems foliowing defo-liation (Andrade & Gomide 1971, Tardin et aI. 1971, Nascimento Júnior & Pinheiro 1975, Gomi-de et ai. 1979, Dovrat et ai. 1980). AndraGomi-de & Go-mide (1971) observed that a rapid and early elongation of intemodes in elephantgrass resulted in the elimination of a higher number of apical meristems. Stem growth, promoted by stem elongation, increases apical meristems vulnerabiity to defoliation (l3ooysen etal. 1963, DahI & Hyder 1977, Gomide et al. 1979, Humphreys 1981). In the Andrade & Gomide (1971) trial ali apical
meri-stems were eliminated when the plants were cut at 56 days of age. Cutting each 28 days allowed the sui-vival of apical meristems which in turn engen-dered good regrowth. Muidoon & Pearson (1979 b) reported that the initial regrowth ofhybrid
Penni-setum was deiayed when the apical meristems were decapitated. When the apical meristems were left intact, a more rapid initial regrowth was observed. The contours ofequal response (Fig. 4, 5, 6 and 7) indicated dlearly that the combined effect of grazing pressure and grazing cycie caused marked changes in the morphology ofdwarfeiephantgrass. Stem height ranged from 13.7 cm (250 kg RLDM Pesq. agropec. bras., Brasília, 21(11):1209-1218, nov. 1986.
ha and continuousgrazing) to 41.4cm (2,500kg RLDM ha' and 56-day cycle). Fig. 4 suggests that stem height was increased as the grazing cycle was increased and the grazing pressure was decreased. At high graxing pressures stem height was gready reduced. Stems were smailer at high grazing pres-sures and short grazing cycies. However, even under continuous grazing, stem growth occurred when the plants were subjected to a more lenient grazing pressure. Ia o —J a, Id
=
(1) (1) uJ 0 o 4 o O 4 25 42 56GRAZING CYCLE (doys)
FIG. 3. Contour map of height of apical meristems of
dwarf elephantgrass as affected by Iength of grazing cycle and grazing pressure. (IR2 0.94, CV =9.16%, P <0.01).
4
The number of axillary buds per tiller was influenced by the sarne conditions which affected stem growth. High grazing pressures, independenily of grazing cycie, depressed the number ofaxi!ary buds per diler (Fig. 5). An exarnination of Fig. 6 and 7 suggests that the appearance ofaxillary buds was associated with the growth of well deveioped stems. The interaction of grazing pressure and grazing cydle is also suggested since the effect of high grazing pressures is to increase the number of axiliary buds at !ong grazing cydles, and iow grazing
pressures appear to compensate for the effect of short grazing cycles.
The number of intemodes per tiller of dwarf elephantgrass varied from 12.3 to 19.8. Greater numbers of internodes were observed at low grazing pressures in all graring cycles (Fig. 6). Although Fig. 6 suggests an increase ia the number of internodes as grazing pressure is decreased and grazing cycle is increased, the '0W variation in
number of intemodes supports the idea that the number of internodes is geneücally established and inlierent to each genotype. Because only visible intemodes were counted it is possible that greater ellniination of apical meristems at higlier grazing pressures prevented the fuil expression of this morphological characteristic.
Cultivars of elephantgrass differ in many mor-phological characterisücs (Bogdan 1977, Muldoon & Pearson 1979 a). However, there is no doubt that differences in height between tal1 and dwarf varieties of elephantgrass are due to the Iength of intemodes. The pattern of internode elongation is due to differenúaüon of celis of apical meristems.
'o o -J ir o, tu a: (/) (1) tu a: o- o z NJ a: o O 14 28 42 56
GRAZING CYCLE (davs)
FUI 4. Contour map of stern height af dwarf
elephant-grass as affected by length of grazing cyole and grazing pressure. (R 2 - 0.95, CV 6.41%, PC 0.01). 'o -c o -J a: o, tu a: (1) (/) tu a: o- o z NJ c
a:
o O 14 28 42 56GRAZING CYCLE (doys)
FIG. 5. Contour map af number of axillary buds per
til-ler of dwarf elephantgrass as affected bï Iength of grazing cycle and grazing pressure. (R - 0.88. CV = 936%, P < 0.01).
Furthermore, elongadon of internodes increases plant height, reduces the density of buds dose to the ground, changes canopy structure, reduces the number of growing points that remain after defo-liation, and may also te associated with a decline in nutritive value (Booysen et al. 1963, Jewiss 1966, Langer 1972, Dali! & Hyder 1977, Gomide et al. 1979, Dovrat etal. 1980, Humphreys 1981). Ia this research, the most interesdng response of dwarf elephantgrass was the reduction of the elongaüon rate of internodes at high grazing pres-sures and short grazing cydles. By maintaining shorter internodes and assuming a more prostrate growth habit the grass quicldy adapted itself to frequent and intense defoliation. Although this phenotypic adaptation was dlearly evident in plants subjected to continuous grazing and the highest grazing pressure (250 k8 RLDM ha'), it was also observed at high grazing pressures ia combination with grazing cycles of 14 and 28 days (Fig. 7). This morphological response prevented
the remova) of some apical meristems which in turn may have allowed herbage production to continue throughout the grazing season. Fig. 6 also suggests that as the grazingpressure was decreased and the grazing cycle was increased, the length of internodes was increased. Another interesting fact is that elongation of internodes was not affected by the length of grazing cyc)e when the plants were subjected to lower grazing pressures (2,500kg RLDM ha'). o o —J a: LiJ a: (1, (1) Lii o: o- o
z
Ni 4 o: (D O 14 28 42 56GRAZING CYCLE (days)
FIG. 6. Cootour map of number of interoodes per tilier of dwarf elephantgrass as affected by Iength of grazing cycle and grazing pressure. IR2 = 0.82.
cv = 6,80%, PC 0.01).
Tiliering of Dwarf Elephantgrass
An analysis of Fig. 8 indicates a decrease in the number of main tiliers per plant as the grazing cycle was shortened and the grazing pressure was increased. The number of main üllers per plant varied from 15.4 (250 kg RLDM hi' and contin-uous grazing) to 34.9 (2,500 kg RLDM ha and 56 day cycie).
lo tropical grasses the number of tillers per plant is usually low when compared with temperate species (Langer 1963, 1972, Jewiss 1966, 1972, Pedreira 1975, Bogdan 1977, Muldoon & Pearson 1979 a, Mott 1983). The number of main tiiers
Pesq. agropec. bras., Brasi'lia, 21(11):1209-1218, nov. 1986.
per plant found in this research is in agreement with the iow number of tiliers found by Boddorff" (1982) in dwarf elephantgrass and in interspecific hybrid of Pennisetum aniericanuni x P. purpureunt.
250 cm o 2001 o -J a: o, 1501 o:
=
o: (001 Oz
N cj o: O 0.7 250 1 I" 1 1, 1 O 14 28 42 56GRAZING CYCLE (dovs)
FIG. 7. Contour map of Iength of internodes of dwart elephantgrass as affected by Iength of graz-
ing cycle and grazing pressure. = 0.94, CV =5.78%,P <0.01).
In contrast to the number of main tiliers per plant, the number of basal dilers per tilier in-creased as the grazing cycle was shortened and the grazing pressure wasincreased (Fig. 9). The number of basal tiliers per main tifler varied from 0.05 (2,500 kg RLDM ha 1 and 56-day cycie) to 0.65 (250 kg RLDM ha' and continuous grazing). The reduction in the length of internodes, as already discussed, and a higher production of basai tiliers resulted in plants with a more prostrate growth ha-bit. h shou)d be emphasized that this morpho)ogical reaction is a clear evidence of the capacity of the grass to adapt itself to stress conditions. In the pasture environment tillering may be significantiy affected by grazing. According to Youngner (1972), when stem apices are removed, apical dominance is broken and the rate oftillerproduc-
.3
don is increased by activation of basal and lateral buds. Defoliation may enhance tiliering, even when apices remained undamaged, by creating a more favorable light environment which would be expected to stimulate the growth and activation of buds (Jewiss 1972, Laude 1972, Youngner 1972). 'o -c o —J 01 tiJ
=
0) (f) w cc oz
N4 cc, o . 0 14 28 42 56GRAZING CYCLE (doys)
FIG. 8. Contour map of numbor of maio tillers por plant
of dwarf elephantgrass as affected by Iength of grazing cycle and grazing pressure. (R2 = 0.60, CV = 14.54%,P <0.01).
The number of primary tiliers per duler of dwarf elephantgrass affected by grazing pressure and grazing cycle is shown in Fig. 10. A greater number of primary tillers were observed at high grazing pressures and short grazing cycles. l-Iowever, at longer grazing cycles the number of primary tiilers tended to increase as the grazing pressure was decreased. This change ia the effect ofgrazing pressure with grazing cycle is an indicadon of a grazing cycle-grazing pressure interaction (Fig. 10). From Fig. 8, '9 and 10, it can be concluded that primary tillers consdtute the predominant form 9f tillering in dwarf elephantgrass under a wide range of management systems. lndeed, a greater number
of primary dilers were found in ali pastures in comparison to basai dilers. This tiliering response is supported by the research of Belyuchenko (1980), who observed that the proportion of new tiliers in the regrowth ofelephantgrass is rnodified by the height of cutdng. When plants were cut at 10 cm or 15 cm, tiliers originated from basal buds of the crown. Cuts at 15 cm or higher produced a higher number of aerial dulers arising from axil-lary buds. 2500 o 2000 o 0.09 —J cc o, 1500 LIJ cc 5 (1) (1) LiJ I00c o
z
N-J 4 250'' O 14 28 42 56GRAZING CYCLE (days)
FIO. 9. Contour map of number of basal tillers por
til-ler of dwarf elephantgrass as affected b length of grazing cycle and grazing pressure. (R = 0.74, CV =39.25%, P <0.01).
The presence of primary tillers ia ali treatment combinadons, even in those with low percentage of apical meristems eliminated, should be consid-ered as an indication of tiliering ability of dwarf eiephantgrass and as a measure of its efficiency as a forage producer.
Secondary tiliers were absent or appeared at a iow rate in pastures subjected to '0w grating pressures. On the other hand, independent of the grazing cycle, high grazing pressure (250 kg RDLM hi') stimuiated the appearance of secondary til-
lers. Increased shadirig, due to the development of greater canopy closures, and low eimination of apical meristems may have induced bud inhibiüon at '0W grazing pressures and long grazing cycles.
40 'o o, - 500 lJJ 2.56 w
,
1000° z
si
3.04\
500 250 3.36 \ O 14 28 42 56GRAZING CYCLE (doys)
FIG. 10. Contour map of primary tiliers por tilier of
dwarf elephantgrass as effected by Iength of
grazing cycle and grazing pressure. (R2 - 0.83,
CV 14.82%,P <0.01).
Dry Matter Distributon in Tiliers of Dwarf Ele-phantgrass
Dry matter production iii tillers was increased by reducing the grazing pressure and increasing the lengtl' of the grazing cycle (Fig. 11). From Fig. 11 it caia be calculated that tiliers ofplants subjected to low grazing pressure and long grazing cycles (2,000 kg and 2,500 kg RLDM ha" and 42 - and 56-day cycles, respectively) were producing approximatety 5 tinies the dry matter when compared with üllers subjected to continuous grazing and high grazing pressure (250 kg RLDM ha"). The magnitude of the response is even higher if we consider that in the highest grazing pressure and short grazing cycle (250 kg RLDM hat and continuous grazing) the number of main
tillers per piant was drastically reduced. There is no doubt that the decrease in DM produced per tilier under intensive grazing was due to the effects of short grazing cycles and high grazing pressures. Therefore, it should be expected that survival of the plants or number of tillers per piant may have affected the DM produced per tilier. From surveys conducted at the beginning and end of the experi-mental period li Was deterrnined that the popu-lation density was about 7,500 piants ha" and the survival of plants vias 100% regardless of treatment. It was also observed that the basal area of plants subjected to continuous grating and higher grazing pressure (250 kg RLDM ha') was drastically reduced, whereas iii pastures subjected to low grazing pressures and long grazing cycles the basal area of the clumps increased. Reductions and increases in the area occupied by the plants may be explained by changes in number and size of main tillers per plant as well as by the variation in number of basal and primary tillers per üller. Furthermore, changes in canopy structure and size of the plants may have affected the competition for light, water, and nutriente, between tillets within plante, and hence the dry matter produced per tiller as indicated by Donald (1963) and Rhodes & Stem (1978).
The effect of the experimental variables upon leaf biade, leaf sheath, and stem DM produced per ten tillers WaS similar to that of total dry matter per ten tillers. Favorable conditions for high DM production of leal biades, leaf sheaths, and stems were met at low grazing pressures in combination with long grazing cycles.
In this experiment, grazing pressure leveis were imposed on the basis of residual leaf biade DM and therefore the increase iii DM produced iii different parte of the tillers as the grazing pressure is de-creased and the grazing cycle is inde-creased may be reiated to the interaction ofseveral factors includ-ing: the residual leaf area left after each grazing period, fluctuations in carbohydrate reserves, and morphoiogical changes of the plants as weli as a greater elimination of apical meristems under high grazing pressures. The response observed is consistent with the increase in growth rate of this grass as reported by Veiga (1983). Since total DM production consists of the contribution of indivi- Pesq. agropec. bras., Brasília, 21(11):1209-1218,nov. 1986.
dual tiliers these resulta are supported by results from cutting experiments (Werner et ai. 1965166, Vieira & Gomide 1968, Andrade & Gomide 1971), and grazing triais where conditions of low grazing pressures and long rest periods after each defolia-tion resulted in increased forage DM production (Maraschin 1975, Serro 1976). 250C 9DM lOti o j20( 297 Lii Ir (1) (1) tu (9 NJ 4 (9 O 14 28 42 56
GRAZING CYCLE (days)
FIG. 11. Cóntour map of total dry marter in tiliers of dwarf elephantgrass as affected by Iength of grazing cycle and grazing pressure. (R2 = 0.96. CV = 8.75%, P <0.01).
CONCLUSIONS
1. Plant height and stem height increased as the grazing cycle was increased and the grazing pres-sure was reduced. The apical meristems remained low at high grazing pressures and short grazing cycies and increased in height as the grazing cycle increased and the grazing pressure decreased. The percentage of apical meristems eliminated varied from 25.00%to 73.33% and was greater under continuous grazing and higher gTazing pressure. The number of axillary buds per tiller was depres-sed under high grazing pressures, independent of
grazing cycle. The number of internodes per tilier tended to increase as the grazing cycle was in-creased and tlie grazing pressure was decreased. The length of internodes was remarkably sensitive to the management imposed. Shorter internodes were observed at short grazing cycles in combi-nation with high grazing pressures.
2. The number of main dilers per plant of dwarf elephantgrass was '0w and decreased from
34.9 to 15.7 as the grazing cycle was shortened and the grazing pressure was increased. Primary tiliers constituted the predominant form of til-lering in dwarf elephantgrass under the wide range ofmanagement systems studied.
3. The morphological responses observed indicate that short grazing cydes and high grazing pressures should be avoided in the management of dwarf elephantgrass in order to insure persistence and productivity of the pasture.
REFERENCES
ANDRADE, I.F. & GOMIDE, J.A. Curva de crescimento e valor nutritivo de capim-elefante (Pennisetum purpureum Schum.) A-146 Taiwan. R. Ceres., IS: 431-47, 1971.
BELYUCHENKO, I.S. Features of regrowth ofpaniculate and eragrostoid perennial grasses. In: WOJAIIN, E. & TIIONS, J., ed. Proceedings of the XIII Internatio-na! Grassland Congress. Berlin, Akademie, 1980. p193-6.
BODDORFF, D. Forage quality of pearl miliet x napier-grass hybtids and dwarf napiernapier-grass. Gainesvilie, Univ. o! Florida, 1982. Tese Mestrado.
BOGDAN, A.V. Tropical pasture and fodder plants. London, Longman Group, 1977, 415p. (Tropical agriculture)
BOOYSEN, P.V.; TAINTON, N.M.; SCOTT, J.D. Shoot--apex dcvelopment in grasses and its importance In grassland management. llerb. Abstt, 33:209-13,
1963.
CASTILLO-GALLEGOS, E. Regiowth and forage quality in Pennisetum purpureum (L.) Schum. var. "dwarf" lime N-75. Gainesville, Univ. of Florida, 1983. Tese Mestrado.
DAHL, B.E. & I-IYDER, D.N. Developmental morphology and management implications. In: SOSEBEE, R.E., ed. Rangeland ptant physiology. Denver, Soc. Range Manage., 1977. p257-90.
DONALD, C.M. Competition among crops and pasture plant&Adv.Agron., 15:1-118,1963.
DOVRAT, A.; DAYAN, E.; KEULEN, H. van. Regrowth potential of shoot and of roots of Rhodes grass (Chioris gayana Kunth.) after defoliation. Neth. J. Agrie. Sei., 28:185-99, 1980.
CARdA, R. & SILVA, U.R. da. Produtividade e carac-terísticas morfológicas do capim-buffel (Cenchrus ciliaris L) cv. gayndah. It. Soe. liras. Zoot., 9:32--42, 1980.
GOMIDE, LA.; OBEID, LA.; RODRIGUES, L.R.A. Fatores morfológicos de rebrota do capim-co1onio (Panicum maximum). R. Soe. liras. Zoot., 8:532-62, 1979.
IIUMPHREYS, L.R. Environmental adaptation ef tropical pasture plants. London, Macmiilan, 1981. 257p. HUMPIIREYS, L.R. Pasture defolialionpractice;areview.
J. Aust. Inst. Agiic. Sd., 32:93-105. 1966.
JEWISS, O.R. Morphological and physiological aspects of growth of grasses during the vegetative phase. In: MILTHORPE, F.L. & IVINS, 1.11, cd. The growth of cereais and grasses. London, Butterworths, 1966. p.39-S6.
JEWISS, O.R. Tiliering in grasses; its significance and contrai.J. Br. Grassi. Soe., 27:65-81, 1972. LANGER, R.H.M. How grasses grow. London, E. Arnold,
1972. 60p.
L&NGER, R.H.M. Tiliering ia herbage grasses. Hcrb. Abstr., 33ç141-8, 1963.
LAUDE, H.M. Extemal factors affecting tilier develop-mcnt In: YOUNGNER. V.B. & MCKELL, C.M., cd. The bioIor and utiliration of grasses. New York, Academic, 1972. p. 146-54 .
LIrrELL, R.C. & MOTT, G.O. Computer assisted design and analysis of response surface experiments ia agronomy. Soil Crop Sd. Soe. Fia. Proe., 34:94-7, 1975.
MARASCUIN, G.E. Response of a coinplex tropical pasture mixture to different grazing management systems. Gainesville • Univ. of Florida, 1975. Tese PhD.
MOfl, G.O. Potential productivity of temperate and tropical grassland systems. In: SMITEI, J.A. & HAYS, V.W., cd. Proceedings of the XIV International Gras-sland Congress. Boulder, Westview. 1983. p.3542.
MULDOON, D.K. & PEARSON, C.J. lhe hybrid between Pennisetum americanum and Pennisetum purpureum. Hcrb. Abstr 49:189-99, 1979 a.
MULDOON, D.K. & PEARSON, C.J. Morphology and physiology of regrowth of the tropical tallgrass hybxid Pennisetum. Ann. Bot.,43:719-28, 1979 b. NASCIMENTO JÚNIOR, D. & PINHEIRO, J.S.
Desenvol-vimento vegetativo da capim jaxagut R. Soe. liras. Zoot.,4:147-67, 1975.
PEDREIRA, J.V.S. Hábitos de perfilhamento do capim colonio, Panicum maximum Jacq. B. Ind,istx. anirit, 32:111-5, 1975.
PEDREIRA, J.V.S. & BOIN, C. Estudo de crescimento do capim elefante variedade Napier (Pennisetuin purpu-reum Schum.) B. lndústr, anim., 26:263-73. 1969. RHODES, 1. & STERN, W.R. Competition for ligilt. In: WILSON, J.R. cd. 1'lant relations ia pastares. Mel-bourne,CSIRO, 1978. p.175-89.
RODRIGUES, L.R.A. Morphciogical and physkilogical responses of dwarf elephantgrass (Fennisetum purpu-reuni (1..) Schum.) to grazing management. Gaines-vilie, Uaiv. of Florida, 1984. Tese Ph.D.
SERRÃO, E.A.S. The use of response surface design ia the agronomic evaluation of a grass-Iegume mixture • under grazing. Gainesville, Univ. aí Florida, 1976.
Tese Ph.D.
TARDIN, A.C.; CALLES, C.H.; GOMIDE, J.A. Desenvol-vimento vegetativo do capim Guatemala. Experiea-• tiae, 12:143, 1971.
VEIGA, J.B. Effect of grazing management upon a dwarf elephantgrass (Pennisetum purpureumSchum.) par-ture. Gainesville, Unif. of Florida, 1983. Tese Ph.D. VIEIRA, L.M. & GOMIDE. JA. Composiçio química e produço forrageira de tr6s variedades de capim elefante. R. Ceres., 15:245-60,1968.
WERNER, J.C.; UMA, F.P.;MARTINELLI, D.;CINTRA, B. Estudo de trds diferentes alturas de corte em capim elefante napier. B. Indóstr. assim., 23:161-8.
1965/66.
YOUNGNER,V.B.Physiclogy ofdefoliationandregrowth. In: YOUNGNER, V.B. & MCKELL. C.M., cd. lhe biology and utilizaüoa of grasses. New York, Aca-demic, 1972. p.292-303.